UMDSub at SemEval-2018 Task 2: Multilingual Emoji Prediction Multi-channel Convolutional Neural Network on Subword Embedding

TL;DR

This paper presents UMDSub, a multilingual emoji prediction system using a multi-channel CNN with subword embeddings, achieving a 2% improvement over character or word-based methods in English tweet emoji prediction.

Contribution

The paper introduces a multi-channel CNN model utilizing subword embeddings for emoji prediction in tweets, enhancing performance over traditional character or word-based approaches.

Findings

Achieved 2% better accuracy than character or word-based models.

Placed 21st out of 48 in SemEval-2018 Task 2.

Demonstrated effectiveness of subword embeddings in emoji prediction.

Abstract

This paper describes the UMDSub system that participated in Task 2 of SemEval-2018. We developed a system that predicts an emoji given the raw text in a English tweet. The system is a Multi-channel Convolutional Neural Network based on subword embeddings for the representation of tweets. This model improves on character or word based methods by about 2\%. Our system placed 21st of 48 participating systems in the official evaluation.